|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | 远洋渔船电子监控技术应用研究进展及展望显示文摘目前,中国正加强远洋渔业的监督和规范管理,长期以来,观察员在远洋渔业管理中发挥重要作用,但其低覆盖率一直是亟待解决的问题,新出现的电子监控(electronic monitoring,EM)成为其有效替代和辅助手段。通过分析国内外远洋渔船EM技术发展,闸明了远洋渔船运用EM的必要性。介绍了EM的组成、具体方法、目标种类及其在远洋渔船常用3种渔具的应用;论述了现今EM系统在远洋渔船的适用性及功能和EM的优缺点:EM技术在远洋渔船已有应用且具有一些观察员无法具备的优势,但缺乏针对各类远洋渔船和目标鱼种监控的高质量应用示范,且对渔业数据相关处理技术的研究精度与深度还不够。最后对EM研究现状进行了总结并对EM技术未来的研究方向和重点提出了具体建议,可为远洋渔船应用电子监控相关研究提供参考。 | 孙永文 张胜茂 蒋科技 樊伟 隋江华 朱文斌 | 2022 | 海洋渔业2022，44，1： | 6 |
| 2 | CNN-LSTM在日本鲭捕捞渔船行为提取中的应用显示文摘目前,国内外部分自动化程度较高的捕捞渔船逐渐开始采用人类观察员与电子观察员相结合的捕捞监控方式。为简化电子监控系统数据处理流程、提高电子监控系统自动化程度、增强相关渔业公司及政府部门的管理效率,该研究提出将卷积神经网络(Convolutional Neural Network, CNN)与长短期记忆模块(Long Short Term Memory, LSTM)应用在传统电子监控系统数据中,用于灯光渔业日本鲭捕捞渔船行为提取。根据日本鲭捕捞渔船工作特点,划分出放网、收网、进鱼、转载等9种捕捞作业行为。设计4组平行试验,分别对比3层卷积神经网络、带有LSTM模块的3层卷积神经网络、8层卷积神经网络、带有LSTM模块的8层卷积神经网络在日本鲭捕捞渔船行为提取中的表现。试验结果显示,在测试集中该4种神经网络训练得到的模型综合评价指标F1分数分别达到了0.794、0.799,0.966及0.972,在配备NVIDIA Tesla V100 32GB高性能GPU的超算环境下5 000组数据平均每次检测耗时分别为34.66、34.50、37.41、37.61 ms。在一定范围内,网络深度的增加会显著提升模型的效果,但检测耗时也会有明显增加;LSTM模块的加入对网络模型效果有一定程度的提升,且不会显著影响检测耗时。因此,CNN-LSTM模块在电子监控系统高实时性、高精度场景下均有较大的应用前景,能够提高电子监控系统自动化程度,提高相关部门的管理效率。 | 王书献 张胜茂 唐峰华 石永闯 范秀梅 樊伟 孙宇 | 2022 | 农业工程学报2022，38，7： | 2 |
| 3 | Advancing multispecies fishery management in China:Lessons from international experience显示文摘China’s 13th Five-Year Plan elevated the national mandate for environmental sustainability.Chinese fisheries are characterized by full retention of high diversity catch harvested using unselective gears,creating ecological risks.Therefore,China launched pilot projects in management by Total Allowable Catch(TAC)in five coastal provinces in 2017 and 2018 to build experience with output controls.Fujian province launched an important pilot in its swimming crab fishery,the first to adopt a multispecies approach.To guide Fujian and other provinces in multispecies management,a workshop in April 2018 shared international experience.The workshop considered 13 case studies spanning a wide range of underlying scientific models and types of harvest controls.Multispecies harvest controls based on simple survey-or index-based models that aggregate trends for many species are typically operationally easier for managers and fishers.However,inadequate management can cause declines of individual species,sometimes leading to adoption of species-specific models and then species-specific harvest controls.This transition often incurs economic costs through scientific and management demands,and constraints on harvest of co-occurring species.The lessons revealed by the case studies suggest multispecies TACs might be effective in the Fujian swimming crab fishery given the modest number of species with similar and productive life history traits,and the market demand for all species.Continued experimentation with different management approaches through pilot projects can enable China to maintain progress toward sustainable fisheries goals under the 14th Five-Year Plan. | Jacob PKritzer Yi Tang Yong Chen Chris Costello Sarah Gaichas Tom Nies Ernesto Penas Keith Sainsbury Changchun Shen Cody Szuwalski Wenbin Zhu | 2023 | Aquaculture and Fisheries2023，8，3： | 0 |
| 4 | Special topic-Fisheries monitoring:Eyes in the ocean:Innovative and multidimensional solutions to monitor fisheries显示文摘In the era of climate change,the visibility of environmental changes dictates public attention.Pictures of untamable bushfires,intense hurricanes,collapsing ice sheets are all gripping images that alarm us and urge us to take action.The ocean,however,gives us less such visuals;the changes that are taking place there are often abstract and hidden.However,the environmental challenges in the ocean are less visible but no less grave;they come from multiple sources:pollution,plastic waste,ocean surface warming,ocean acidification(IPCC,2019).But the most direct and age-old impact we exert on the ocean and its ecosystems is overfishing(FAO,2018). | Bowen Chang Shu Su Chongliang Zhang Yiping Ren Xiujuan Shan Jake Kritzer Yong Chen | 2020 | Aquaculture and Fisheries2020，5，3： | 0 |
您所在的机构暂未开通此部分的功能，请联系管理员